If $f(x) = 0.5 e^{-|x|}$ for $-\infty < x < \infty$, how would you find the moment generating function for this? Also how would you find the distribution of $Y = |X|$?


$$E(e^{tX}) = \int_{-\infty}^\infty f(x) e^{tx} \; dx.$$

  • 1
    $\begingroup$ Hint: continue your attempt, separating the integral into an integral on $x\lt0$, where $f(x)=\frac12\mathrm e^x$, and an integral on $x\gt0$, where $f(x)=\frac12\mathrm e^{-x}$. $\endgroup$ – Did Feb 28 '12 at 13:52
  • $\begingroup$ I get (1/(t-1))*(e^infinity -1) when x is positive. I am not sure what I am doing wrong. $\endgroup$ – lord12 Feb 28 '12 at 14:11
  • $\begingroup$ The pdf doesn't integrate to one. I think there is a .5 in the exponential term. $\endgroup$ – Stuck_pls_help Feb 28 '12 at 15:01

As Didier Piau stated: $$ \begin{align} E(e^{tX}) & = \int_{-\infty}^\infty f(x) e^{tx} \ dx=\int_{-\infty}^0 f(x) e^{tx}\ dx+\int_0^\infty f(x) e^{tx}\ dx \\ \\ & =\int_{-\infty}^0 0.5e^x e^{tx}\ dx + \int_0^\infty 0.5 e^{-x} e^{tx}\ dx \\ \\ & = 0.5\left(\int_{-\infty}^0 e^{(t+1)x}\ dx+\int_{0}^\infty e^{(t-1)x} \ dx\right) = \frac{0.5}{t+1}-\frac{0.5}{t-1},\quad t\in (-1,1) \end{align} $$

For the distribution of $Y=|X|$ we have the general rule for a transformation of the form $y=h(x)$ in our case $Y=|X|$ with $Y\in [0,+\infty)$: $g(y)=f(h^{-1}(y))|\frac{d(h^{-1}(y))}{dy}|$. Applying this rule to your problem will give you the distribution of Y.

  • 2
    $\begingroup$ One can simplify $E(\mathrm e^{tX})$ to $1/(1-t^2)$. $\endgroup$ – Did Feb 28 '12 at 15:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.